Subscriber access provided by The Chinese University of Hong Kong
Article
Bicarbonate and Alkyl Carbonate Radicals: Their Structural Integrity and Reactions with Lipid Components Michael Buehl, Peter DaBell, David W. Manley, Rory P. McCaughan, and John C. Walton J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.5b10693 • Publication Date (Web): 01 Dec 2015 Downloaded from http://pubs.acs.org on December 1, 2015
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 14
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of the American Chemical Society
Bicarbonate and Alkyl Carbonate Radicals: Their Structural Integrity and Reactions with Lipid Components.
Michael Bühl,* Peter DaBell, David W. Manley,† Rory P. McCaughan and John C. Walton.* University of St. Andrews, EaStCHEM School of Chemistry, St. Andrews, Fife, KY16 9ST, UK Supporting information
ABSTRACT: The elusive neutral bicarbonate radical and the carbonate radical anion form an acid/conjugate base pair. We now report experimental studies for a model of bicarbonate radical, namely methyl carbonate (methoxycarbonyloxyl) radical, complemented by DFT computations at the CAM-B3LYP level applied to the bicarbonate radical itself. Methyl carbonate radicals were generated by UV irradiation of oxime carbonate precursors. Kinetic EPR was employed to measure rate constants and Arrhenius parameters for their dissociation to CO2 and methoxyl radicals. With oleate and cholesterol lipid components methyl carbonate radicals preferentially added to their double bonds; with linoleate and linolenate substrates abstraction of the bis-allylic H-atoms competed with addition. This contrasts with the behavior of ROS such as hydroxyl radicals that selectively abstract allylic and/or bis-allylic H-atoms. The thermodynamic and activation parameters for bicarbonate radical dissociation, obtained from DFT computations, predicted it would indeed have substantial lifetime in gas and non-polar solvents. The acidity of bicarbonate radicals was also examined by DFT methods. A noteworthy linear relationship was discovered between the known pKas of strong acids and the computed numbers of microsolvating water molecules needed to bring about their ionization. DFT computations with bicarbonate radicals, solvated with up to 8 water molecules, predicted that only 5 water molecules were needed to bring about its complete ionization. On comparing with the correlation, this indicated a pKa of about -2 units. This marks the bicarbonate radical as the strongest known carboxylic acid.
INTRODUCTION Carbonic acid is a weak, diprotic acid formed upon dissolution of carbon dioxide in water. While its acid/base equilibria are well understood, its radical chemistry is not. Deprotonations produce the bicarbonate anion HCO3− (pKa = 6.38) and then carbonate dianion CO32− (pKa = 10.25) (See Scheme 1).
Scheme 1. Formation of Carbonic Acid and Associated Anions and Radicals.
These species play important roles in geology, oceanology, atmospheric chemistry and of course in physiology. In blood serum and intracellular media these equilibria constitute the ‘bicarbonate buffer system’. This is critical for maintaining the pH constant within the range 7.35-7.45 which is essential for optimum functioning of enzymes.1 Approximately 70 % of CO2 is transported as bicarbonate in the human body (25.0 mM in serum and 14.4 mM in intracellular media).2 When an electron is removed from bicarbonate or carbonate respectively, neutral bicarbonate radicals 1 or carbonate radical anions 2 are created. There are several enzymatic (and possibly nonenzymatic) ways these radicals can be produced in biological fluids. For example CO2 is one of peroxynitrite’s primary biological targets producing NO2 and 2.3 Xanthine oxidase turnover of acetaldehyde and other substrates is also known to produce 2.4 Similarly it is recognized5 that the Cu,Znsuperoxide dismutase/H2O2 system also generates 2.
As shown in Scheme 1, the carbonate radical anion 2 forms a conjugate base/acid pair with the neutral bicarbonate radical 1. Apart from attempts to determine its pKa, virtually nothing is known about the chemistry or biochemistry of the neutral bicarbonate radical itself. It was shown to be a strong acid,6 and a computational study7 suggested its pKa might be as low as -4 units. Because of this acidity it is generally assumed that carbonate radical anions 2 are the dominant partner of the pair in biological fluids. Carbonate radical anion 2 is a reactive oxygen species (ROS) that contributes to oxidative stress.8 It is an important oxidizing agent in aqueous solution,9 and its main bio-targets are polar species such as biothiols, nucleic acids, metalloproteins/proteins and glutathione.10 Because HOCO2• is neutral, whereas −OCO2• is negatively charged, and the two radicals have very different extents of electron delocalization, their preferred reaction channels and reactivity are expected to be markedly different. Thus 2 is lipophobic, with the majority of its physiological reactions taking place in polar environments. Therefore, it should be a poor initiator of lipid peroxidation, due to its low diffusibility in hydrophobic environments. The neutral protonated radical 1 could well be the dominant form in non-polar hydrophobic lipid rich situations. Initiation of peroxidation is the likely role of 1 in structures such as membranes, vesicles, lowdensity lipo-protein particles (LDL) or lipid microdomains. However, as of yet, there is no research reporting peroxidation of lipids by 1. One reason for this is that the carbonate and bicarbonate precursors of 1 are only soluble in polar solvents so that even when 1 is formed it immediately deprotonates to 2. No lipid soluble precursor for 1 is currently known. To understand oxidative lipid and cell damage, with the resultant functional decline and associated degenerative conditions, a thorough study of the chemistry and biochemistry of both components of the carbonate/bicarbonate radical pair is much needed. Radical 1 may be viewed as the first member of a homologous series with alkyl carbonate radicals: HOC(O)O, MeOC(O)O, CnH2n-1OC(O)O
ACS Paragon Plus Environment
Journal of the American Chemical Society
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
These neutral alkyl carbonate radicals are structurally similar to 1 so they, and particularly the methyl carbonate radical (7), are suitable models for 1. Several members of this series have been generated and precursors soluble in organic solvents, that is dialkyl peroxydicarbonates11 and oxime carbonates,12 are known. However, to our knowledge, the simplest member, methyl carbonate, has not yet been reported. The objectives of this research were first to weigh up lipidsoluble precursors for radical 1. Second to investigate the structural integrity of 1, and of model MeOC(O)O• and related species, particularly in respect of the ease with which they decarboxylate: R1OC(O)O• → R1O• + CO2 An examination of specific reactions of these species with lipid components was also a priority. In each case both experimental and quantum-mechanical (QM) computational methods were employed. By these means insight into the reactivity of model MeOC(O)O•, and of radicals 1 and 2, as well as new insights into the pKa of 1 was obtained. Understanding of the way 1 and 2 differ from hydroxyl radicals - the archetype ROS - in the type of reaction they undergo, and in their site selection with unsaturated fatty acids and cholesterol, was also obtained.
EXPERIMENTAL STUDIES Bicarbonate Radical (1) Precursors. Research with oxime carbonates 3 showed they release alkyl carbonate radicals ROC(O)O• on photolysis. These compounds are safe and have long shelf-lives and are thus promising precursors for the study of HOC(O)O• and MeOC(O)O• radicals. Compound PhCMeC=NOC(O)OH (4) would be expected to release radical 1 on irradiation together with the much less reactive13,14,15 iminyl radical PhMeC=N• (Im) (Scheme 2).
Page 2 of 14
were purged with N2 and UV irradiated in the resonant cavity of an EPR spectrometer. Neither MeOC(O)O• radicals nor MeO• radicals will be directly detectable by EPR in solution.11a,17 The spectra actually disclosed the corresponding iminyl radical together with a second species having g = 2.0026, a(1H) = 34.5, a(1H) = 13.1, a(1H) = 9.3, a(1H) = 8.1, a(1H) = 2.8 G. By analogy with literature data for other alkyl carbonate radicals11,12 we identify this as the meta-adduct radical 8 from addition of 7 to the solvent (see Scheme 3).
Scheme 3. Generation and Reactions of Methyl Carbonate Radicals 7. R Me
O
O
N
UV, MAP Ph
O 6a; R=H b; R=Me
O Im
+
MeO O
PhBu-t 7
PhBu-t
ka
kd MeO
Bu-t
+
CO2 8
OC(O)OMe H
The concentration ratio [8]/[Im] was measured by simulation of spectra and is shown as a function of temperature in Figure 1.
Scheme 2. Potential Lipid Soluble Precursors for Radical 1
Ph
N
deprotection Ph
N
O
O O
3
O O
4
(a)
OH UV -Im O
Z
HO (b) UV -Im
O
O O 5
Z
O rearr. elim.
1
Deprotection protocols with various derivatives of 3 [(Z = protecting group, route (a)] have not so far been successful; nor have experiments to generate radicals 5 that rearrange and/or eliminate to release 1 [route (b) in Scheme 2].16 Before embarking on a project to further develop route (b) we decided to first forecast the structural integrity of 1, and to see how its reactivity would differ from 2 and from HO• radicals, by examination of model MeOC(O)O• radicals and by DFT computations.
Methyl Carbonate Radical Additions to Aromatics and Decarboxylation Kinetics. The oxime carbonates 6a,b were clear choices as precursors for the methyl carbonate (methoxycarbonyloxyl) radical 7. They were prepared by the literature method12a from methyl chloroformate and the appropriate oxime. Solutions of each of 6a and 6b (0.1 mol) in t-butylbenzene containing 4methoxyacetophenone (MAP, 1 equiv.) as photosensitizer
Figure 1. Data for Decarboxylation of MeOC(O)O• Radicals. Filled circles: data from oxime carbonate 6a Open circles: data from oxime carbonate 6b
With the 6b precursor (open circles) the [8]/[Im] ratios were more scattered and less reliable because of the large difference in line-width between 8 and Im in this case and the greater extent of overlap of their spectra. We therefore gave more weight to the data from 6a. Equal numbers of Im and 7 radicals were formed in each initiation step and so the fact that the [8]/[Im] ratio was close to 1 at low temperatures (Figure 1, closed circles) indicated that the meta-addition of radical 7 to the solvent was very fast and complete. It follows, therefore, that the amount of 8 equals the amount of 7 in solution. Above T ~ 250 K the [8]/[Im] ratio decreased and we attribute this to the onset of decarboxylative dissociation of 7 to MeO• and CO2. The rate constants of dissociation of 7 (kd) were determined from measurements of the concentrations of 8 and Im in the
ACS Paragon Plus Environment
Page 3 of 14
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of the American Chemical Society
fall-off region. Making the Steady-State Approximation and assuming all termination steps were diffusion controlled with rate constant 2kt we obtained equ. (1):18 kd/2kt = [MeO] + [MeO]2/[7] (1) Provided other reactions of 7 were insignificant [7] = [8] and hence: kd/2kt = {[Im]-[8]} + {[Im]-[8]}2/[8] (2) 19 Iminyl radicals terminate at the diffusion rate; as do the other radicals in Scheme 4, and hence the use of Fischer’s data for 2kt [logAt = 11.63 M-1 s-1, Et = 2.25 kcal mol-1],20 appropriately corrected for the difference in solvent viscosity,21 was justified. Radical concentrations were determined from double integrations of the EPR spectra derived from precursor 6a in the fall-off region and a satisfactory Arrhenius plot was obtained (see Supporting Information) with parameters as follows: log(Ad/s-1) = 13.9±1.8, Ed = 14.6±2.2 kcal mol-1 kd(300 K) = 1.8 × 103 s-1 These results seem very reasonable because the measured activation barrier is close to experimental and to DFT computed barriers for dissociation of PhCH2OC(O)O• radicals (12.9±2.0 kcal mol-1).12a It is evident therefore that 7 has sufficient structural integrity to take part in a range of chemical processes. Rate constants for H-atom abstraction and addition reactions of alkyl carbonate radicals are of the order of11b 107 to 108 M-1 s-1. Thus radical 7 with a kd of 1.8 ×103 will easily be persistent enough to engage in these processes at 300 K for substrate concentrations of > 10-4 M. There are interesting implications of these findings for the chemistry of bicarbonate radicals 1. These will dissociate by β-scission to CO2 and HO• radicals. The activation barriers and rates of radical β-scission reactions usually depend strongly on the thermodynamic stabilization of the released radical.22 The HO• radical is less thermodynamically stabilized than MeO• or EtO• by 13.6 and 13.4 kcal mol-1 respectively.23 This gives a strong indication that the barrier to decarboxylative β-scission of the HOC(O)O• radical 1 will be considerably higher than that of 7 and the lifetime of 1 in non-polar environments will be even greater. DFT computations,12 (see below) also predict a higher barrier. Once formed, therefore, bicarbonate radicals are expected to have ample lifetimes to attack lipid components and contribute, with other ROS, to their oxidative transformations. Our next step was to examine model reactions of radical 7 with organic substrates and lipid components so as to shed light on how bicarbonate radical oxidative processes could differ from those of HO• radicals and other ROS. As shown above, addition of 7 to aromatics was rapid. We also found that 7 added very efficiently to furan, thiophene and derivatives thereof (Table 1; see Supporting Information for sample spectra). The addition was selective for the 2-(or 5)-positions and both 2- and 5-adduct radical isomers (9) were obtained in approximately equal amounts from 3methylthiophene. In no case was an adduct radical from attack at a 3-position observed although, because of the considerable noise levels, minor amounts would escape detection.
Table 1. EPR Parameters of Adduct Radicals (9) of Methyl Carbonate Radicals and Heterocycles.*
X, R
T/K g-factor
a(H2)
a(H3)
a(H4)
a(H5)
O, H
240
2.0029
19.5
13.6
1.9
14.4
O, 2-Me
290
2.0022
18.4
13.9
1.6
12.7(3H)
O, 2-t-Bu
230
2.0030
19.7
13.9
1.7
-
S, H
240
2.0046
18.6
11.9
2.3
13.7
S, 3-Me§
230
2.0040
17.4
12.5(3H) 2.5
13.7
S, 3-Me§
230
2.0040
18.1
11.2
14.0
* EPR hyperfine splitting (hfs) in Gauss.
2.4(3H) §
Isomeric radicals
from addition at C-2 and C-5. It is noteworthy that addition of 7 to the rings was preferred to abstraction of the ‘benzyl-like’ H-atoms of the CH3 groups attached to both heterocycle types. Reaction with toluene was also examined. The spectra were too weak for definitive analysis but appeared to show a mixture of ortho-, meta- and para-adduct radicals with again no sign of benzyl radicals from H-abstraction by the MOC(O)O• radicals. GC-MS analysis of the photolysate supported this conclusion (see Supporting Information).
Methyl Carbonate Radical Reactions with Lipid Components. Peroxidation, and the associated oxidative stress in organisms, is initiated when ROS damage lipid components of cells.24 The mechanisms associated with this peroxidation have been intensively studied over many years.25,26 The hydroxyl radical is the most reactive ROS and it is well established that this species initiates much peroxidation. It abstracts H-atoms from allylic sites in mono-unsaturated lipid components such as oleic acid and cholesterol and from bisallylic sites in di- and poly-unsaturated fatty acids (PUFA). Subsequent chain propagation proceeds through addition of oxygen to the C-centered radicals, so generating peroxyl radicals that then abstract H-atoms, thus producing hydroperoxides. Our aim was to establish if bicarbonate radicals 1 would initiate in this same way, simply augmenting the regular peroxidation process, or if they could initiate alternative oxidative sequences ending in novel metabolites. Our first evidence that radicals 7 behave differently came from a study of their reaction with hex-1-ene. When a solution of oxime carbonate 6b and MAP in neat hex-1-ene was UV irradiated in the EPR cavity the spectrum showed the PhMeC=N• radical (73 %) and another radical (27 %) with EPR parameters: g = 2.0027; a(1H) = 21.0, a(4H) = 24.7 G at 290 K. We identify this as the adduct radical CH3(CH2)3CH•CH2OC(O)OMe. Surprisingly, none of the allylic radical from H-abstraction adjacent to the double bond could be detected. The predominant process was addition to the C=C double bond. As a control experiment we examined the reaction of 7 with the fully saturated fatty acid derivative methyl stearate [nC17H35C(O)OMe]. The EPR spectrum taken during photolysis of a benzene solution (0.1 M in 6b and 0.16 M in methyl stearate) at 292 K showed Im (87 %) and a minor amount (13 %) of a radical with g = 2.0028, a(1H) = 21.3, a(4H) = 24.7 G. This is clearly a composite spectrum of the secondary radicals –CH2CH•CH2- (s•) formed on H-abstraction from all 14 of the methylene groups in the chain that are flanked on both sides by CH2 groups. This very weak spectrum contrasts with that reported for H-atom abstraction from stearic acid by t-BuO• radicals (a model for ROS).27 In that case strong spectra of s• plus signals from the radicals α-to the COOH
ACS Paragon Plus Environment
Journal of the American Chemical Society
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
group (C2,α-radicals) and adjacent to the CH3 group (C17, ω-1 radicals) were observed and indicated that t-BuO• radicals abstracted rapidly and unselectively from every site. The EPR spectrum obtained from UV photolysis of a similar solution containing 6b, methyl oleate (10) and MAP in PhH at 292 K is shown in Figure 2 (top). In addition to Im radicals, a second radical with g = 2.0025, a(2H) = 26.1, a(1H) = 21.5, a(1H) = 13.8 G was observed. This was readily identified as the adduct radical (11a,b) of MeOC(O)O• to the double bond of the methyl oleate (Scheme 4). A minor amount of secondary radicals s• (g = 2.0025, a(1H) = 21.5, a(4H) = 24.5 G) was also observed. Spectra were obtained at several temperatures down to 220 K (with toluene as solvent for lower temperatures). Small amounts of peroxyl radicals (p, Figure 2, bottom) were detected with this solvent. The relative concentrations of the radicals determined at each temperature were as shown in Table 2. None of allylic type radical 12 from H-abstraction adjacent to the double bond (at C8 or C11) was detected.28 The relatively large amounts of 11 showed that addition of radical 7 was rapid even at 220 K.
Page 4 of 14
sociation of MeOC(O)O• did not compete. The small proportion of s• observed at 292 K demonstrated MeOC(O)O• was less selective for addition at higher temperatures such that some H-atom abstraction occurred. The two radicals 11a and 11b formed on addition of MeOC(O)O• to either end of the double bond of 10 have extremely similar structures in the vicinity of their radical centres. Their EPR spectra will be indistinguishable and the experimental spectrum is probably a 50:50 mixture of the two.29 The 13.8 G doublet hyperfine splitting (hfs) is from the single Hβ of 11 and its small magnitude (and the fact that this decreased to 12.5 G at 220 K) indicates this Hβ lies close to the nodal plane of the p-orbital containing the unpaired electron. We can conclude that the preferred conformation is as shown in 11as.30 The majority of initiation in oleate peroxidation by other ROS types occurs by H-atom abstraction adjacent to the double bond. For t-BuO• radicals with oleate, for example, the allylic type species 12 (plus minor s•) were the only radicals detected by EPR spectroscopy.31 There is therefore a striking contrast between the preferred addition reaction of MeOC(O)O• radicals to C9 and C10 of oleate and H-abstraction from C8 and C11 preferred by other ROS.
Scheme 4. Reaction of Methyl Carbonate Radical with Methyl Oleate. MeOC(O)O E
11
18
10
11a
E
9 8
2
10, E = CO2Me
E
MeOC(O)O 11b MeOC(O)O Hβ
Hα
E
Figure 2. EPR Spectra from 6b and methyl oleate in solution. Top: experiment at 292 K in PhH Centre: Computer simulation; “s” indicates prominent peaks of secondary radicals, ‘Im’’ indicates PhMeC=N• radicals. Below: experiment at 220 K in PhMe; ‘p’ indicates peroxyl radicals.
Table 2. Relative concentrations of radicals obtained on photolysis of 6b and methyl oleate. T/K
Solvent
Im (%)
11a,b (%)
s• (%)
292
PhH
60
36
4
273
PhH
51
49
